What is your most needed feature?
For me it’s table 1 has columns an and b, table 2 has columns b, c and d. if you filter on a in table 1 then table 2 filters due to the matching of b on b. This is fairly normal, but doesn’t universally work in QS
1 Like
Hi @charris - Welcome to “QuickSight Community”. To understand your query very details, can you please share sample input and sample expected output, possibly with some dummy data.
QuickSight has lots of good and insightful capabilities, we all need to see and find a way how we can achieve it. If something is not there, we have a fantastic team from product team who can help in put it as a feature request.
Tagging QuickSight experts
@Max @Thomas @Tatyana_Yakushev @Bhasi_Mehta @Kristin @David_Wong @Biswajit_1993
1 Like
Hi @charris,
Welcome to the community!
The described behavior can be achieved in QuickSight by scoping your filter. Filters can be scoped to a single visual, some visuals, all visuals that use this dataset, or all applicable visuals (meaning the filter applies to any visuals that have valid column mappings).
For more information on this subject, please have a look at step 5 of the Adding filters in analyses section in the documentation.
In your particular example, you would need to scope the filter that filters on column a to filter on the visuals for table 1 and table 2, or all visuals that use this dataset if your tables leverage the same dataset. If you use different datasets for the two tables, you need to scope the filter to all applicable visuals using a column mapping.
I hope my solution worked for you. Let us know if this is resolved. And if it is, please help the community by marking his answer as a “Solution”.
2 Likes
I still think @charris has a point here. Whilst it’s technically perfectly possible in PowerBI, I don’t think it’s well done. You have to learn how to do it (it’s not very obvious), and I still find it easy to make a mistake. Although I will say in QS’s favour, it is highly configurable.
Having come from PowerBI, Microsoft do this waaaaaay better:
It is beautifully simple in PowerBI, very easy to use and very effective. AWS would do well to learn (copy!) from MS on this one.
Thanks for the responses.
I do get how the various filters work (I think!). What I think is missing is this:
- There are 2 datasets with a common column
- Filter the first data set on a different column (not common), then the second one is filtered due to the common column linkage
Well it does not seem to work for me! I need to join the ‘not common’ column into the second dataset before the filtering works.
Hope my explanation is now clearer.
What you’re describing isn’t supported.
Dataset 1: Fields A and B
Dataset 2: Fields B and C
What you can do in QuickSight is filter visuals created from the 2 datasets by their common field B.
What you’re looking for is filter dataset 1 by field A and filter dataset 2 by field B at the same time because both datasets have field B in common.
1 Like
Thanks, that is a very clean description. Yes, I think that is the most needed feature, I have loads of use-cases for it. Are there any others out there thinking the same?
@charris Agreed that this would be very useful. I believe the reason this can be done in Power BI and Tableau because they allow you to create relationships.
FYI there’s another discussion about this particular topic at How to filter multiple datasets using a single filter originating from a separate dataset in Analysis? - #4 by Koushik_Muthanna
1 Like
Thanks for the update, there is some useful information in that link.
I do find myself doing many extra joins to achieve the functionality I need. Also, it’s a very common user question : ‘why doesn’t it filter on that table?’
If Quicksight built the relationship by default (which could be turned off) it would be best, also I think the end users would understand and accept that more readily!